A Joint Handoff and Offloading Decision Algorithm for Mobile Edge Computing (MEC)
A Joint Handoff and Offloading Decision Algorithm for Mobile Edge Computing (MEC)
复制标题
DOI：
10.1109/globecom38437.2019.9013932
复制
发表时间：
2019-12
期刊：
影响因子：
--
通讯作者：
W. Nasrin;Jiang Xie
中科院分区：
文献类型：
--
作者：
W. Nasrin;Jiang Xie
移动边缘计算(MEC)的主要特征是将计算资源和存储推送到网络边缘(例如蜂窝基站(BSS))，以便从移动设备上卸载计算密集型和延迟敏感型应用。然而，部署具有BS的MEC会增加本已拥挤的蜂窝网络的额外流量，这可能会导致无线电拥塞。另一方面，用户移动性不仅会触发两个BS之间的切换，还会触发两个MEC之间的切换，这就给具有足够计算资源的MEC向目标BS执行切换带来了新的独特挑战。遗憾的是，现有的用于传统蜂窝网络的HO判决算法不能解决这些挑战。此外，无线电网络拥堵问题正被完全忽视。本文提出了一种适用于无线和计算卸载的MEC系统的HO判决算法。提出的HO判决算法与毫微微小区协作，通过卸载静态流量来减少无线拥塞的影响。它还与远程云合作，通过卸载延迟容忍计算来减少MEC的拥塞影响。仿真结果表明，本文提出的HO决策算法显著提高了服务的分流成功率和服务失败率。据我们所知，这是第一个HO判决算法，同时解决了MEC系统中的无线电和计算卸载问题以及移动性问题。
The main feature of Mobile Edge Computing (MEC) is to push computational resources and storage to the network edges (e.g., cellular base stations (BSs)) in order to offload computation-intensive and latency-sensitive applications from mobile devices. However, the deployment of MEC with a BS adds extra traffic to the already crowded cellular network that may lead to a radio congestion. On the other hand, user mobility triggers handoffs (HOs) not only between two BSs but also between two MECs, which creates new unique challenges of performing HOs to a target BS with an MEC which must have sufficient computational resources. Unfortunately, existing HO decision algorithms for traditional cellular networks cannot ad- dress these challenges. In addition, the radio network congestion issue is being utterly ignored. In this paper, we propose a novel HO decision algorithm for MEC systems that cooperates both radio and computation offloading. The proposed HO decision algorithm cooperates with femtocells to reduce the effect of radio congestion by offloading static traffic. It also cooperates with the remote cloud to reduce the impact of congestion at the MEC by offloading delay- tolerant computation. Simulation results show that our proposed HO decision algorithm significantly improves the offloading success and the service failure rates. To the best of our knowledge, this is the first HO decision algorithm that addresses both radio and computation offloading issues and mobility together in MEC systems.